Again, IT'S NOT THE SAME DATA, split it in two columns as basic design rules tell us to do.

check this out for more information on this method and why you shouldn't use it :

http://www.sqlservercentral.com/forums/shwmessage.aspx?forumid=8&messageid=189202

Also this may run faster if attribute_id is indexed :

declare @v1 int

set...

You can't update a text column like this, while I can't show you how I can tell you to go link into the books online for more information.

HTH.

Thanx. Looks like an index seek is just out of the question here, so the simplest conversion wins .

I still need the sample data from the base table to rebuild the correct query.

This would denormalize the db, you simply need to select the data correctly when presenting it using inner joins. There's no need the ship all this data in the...

I thaught the other experts would provide more help than that .

"And, since we don't have function indexes in SQL Server as in Oracle, it would be a rather unpleasant thing..."

This isn't 100% accurate :

IF Object_id('Employes2') > 0

DROP TABLE Employes2

GO

IF...

it's a good idea but it doesn't seem to work :

Select

cast('2005-05-02 14:56:52.493' as smalldatetime)

, cast ('2005-05-02 14:57:03.447' as smalldatetime)

, convert(varchar(16),'2005-05-02 14:56:52.493',20)

, convert(varchar(16),'2005-05-02 14:57:03.447',20)

2005-05-02 14:57:00

2005-05-02 14:57:00

2005-05-02 14:56

2005-05-02 14:57

You'll be better off in .net. The string manipulation is much easier there.

I'll have to use a lifeline on that one .

As long as it's fixed now... Happy to almost help .

Can you post some sample data, the current results and the expected results. It will be hard to help you without that.

From the top of my head : Insert OFF HOURS in small batches of 500/1000 reports at the time... Should be done before the next day